A recent blog by David Ashton “Keeping an eye on the International Space Station” posed the question: “There are plenty of embedded systems up there keeping it going – have any readers been concerned with those?” Actually I have been involved. I worked on the Canadarm2, the robotic arm that was Canada’s contribution to the International Space Station. Instead of attempting a lengthy comment to David, I made a short comment and happily turned the rest into this blog. Let me hasten to add that when I worked as a rocket scientist, I in fact was writing specifications – the robotic arm was being defined on paper. I left (unwillingly) before anything real existed and the details that I present may well have evolved and changed since I left.
NASA (do I need to decrypt that?) is a world of acronyms; there are even nested acronyms. It was a major headache for me just to get up to speed as to what they were talking about. I promise to try and define every acronym, although I must warn you that sometimes the real name is just as meaningless as the acronym. For example an “EE” is an “end effector”. Are you any the wiser?
The Canadarm2 (technically the SSRMS – Space Station Remote Manipulator System) was a logical contribution to the proposed space station after Canada had made the Canadarm (SRMS – Shuttle Remote Manipulator System) for the shuttle program. Aside from dimensional details the only real difference was the number of joints – the so-called degrees of freedom. The SSRMS has seven, the SRMS has six. Today Japan and the Europeans also have robotic arms on the ISS (International Space Station), but the SSRMS was the first and was used in the later stages of the original construction. Canada also has produced Dexter (technically SPDM –Special Purpose Dextrous Manipulator), a hand, if you will, to go on the end of the arm.
An artist’s concept of the SSRMS in operation. This signed picture was my parting memento when I left.
It is more important for devices in space to be reliable and safe than to be on the technological cutting edge. Up there you are a long way from home in a very hostile environment. In addition development and space qualification of parts takes so long that the latest technology today is rather dated when it gets up there. Anything that is designed must work in temperature extremes with sudden swings between the extremes, intense vibration, high acceleration and deceleration (deceleration only on the SRMS), and zero air pressure, but there are additional restrictions placed by NASA. For instance the materials used cannot off-gas and there cannot be any metal particles since they become micro-meteorites. This has impacts down to the bolts and nuts – if the fit during assembly up there isn’t perfect you could get shavings coming off.
The SSRMS (and SRMS) are made of carbon fibre and surprisingly cannot support their own weight on Earth. That makes for a problem in testing it. Spar (the original organization that designed it) came up with the solution of mounting on an air cushion (like a hovercraft) at each joint and testing it in one plane at a time. (The proof of concept used a vacuum cleaner as the compressor and the guy who came up with the idea – he wasn’t very popular – earned the sobriquet “Captain Hoover”.) Now consider that the SSRMS had to capture all 100 tons of the shuttle and manoeuvre it around. Everything has to happen sl-o-o-o-o-o-wly, not only because of the mechanical loads on the arm, but also not to impart unwanted movements to two unfixed objects just floating in space.
At each end of the SSRMS is an EE and it mates with a PDGF (Power Data Grapple Fixture) mounted on whatever you want to grab with the arm. The mechanical part of the PDGF looks a bit like an elongated mushroom (we will get to why in a minute) and there are optical targets along with the connectors for power and data. Together with the EEs are cameras that are used to zero in on the target on the PDGF. The mushroom part of the PDGF goes into the EE at capture, but as you mechanically connect to the EE you cannot impart any torque since one or both of the weightless bodies may start to spin and then correcting it is a real problem. The solution was patented which you can see here but in essence the EE is a short hollow cylinder and within it are three wires arranged to form a triangle. The mushroom part of the PDGF (actually the patent is for the PGF – the shuttle’s PDGF equivalent) is aimed for the centroid of the triangle and when there, motors wind the wires so that the triangle gradually gets smaller and the head of the mushroom is captured. Neat, huh? Another fun fact: There are PDGFs along the length of the truss and the SSRMS can be walked from PDGF to PDGF (remember there are EEs at each end of the SSRMS) down the truss if relocation is needed.
I did hear about a new idea where the capture would be done magnetically – perhaps that it what they use now.
In order to discuss the operation of the SSRMS it was always likened to an “arm” and we would try to explain the actions by moving our arms and bending elbows and wrists. It was not really satisfactory since human joints are completely different to the mechanical ones on the SSRMS and there are far fewer of them. A wag actually proposed that we refer to it as a “leg” simply to see the contortions that people might try. One engineer had the bright idea of putting together a model made out of Lego. It worked well and I still have pieces of Lego running around the bottom of a drawer somewhere.
The Lego version of the SSRMS. The insulating cover for thermal protection (in white with “Canada” printed on it) is just paper wrapped around the plastic shaft.
My part in all of this was to maintain and write the specifications of different parts of the MSS (Mobile Servicing System) which is what the whole system including the software on the ISS’s computers was called. At times I worked on the MBS (Mobile Base System) which was then designed out, the SSRMS, the CEI (Contract End Item), the overall contract, and then the only meaningful part, which may still endure. I worked on the transaction analysis of downloading software to the joint computers on start up.
I didn’t mean to make this job sound glamourous, and if you got that impression, I apologise. I spent most of my two years there pushing paper and getting frustrated with the bureaucracy of large projects. Sure there were interesting moments, when NASA brought back a probe (I forget the name) with micro-meteorite damage, going into the test-bed clean room, visiting the SRMS simulator (not a complete shuttle, but close enough) and meeting a couple of astronauts. But the only real achievement I feel was was that I was a real whizz at Wordperfect when I left. Oh, and I also got the seeds of a concept that became a very popular Design Idea on using Excel to create a Traceability Matrix. For more on that see the first comment in this embedded.com blog “Software Standards Compliance 101: Using a formal requirements capture process.” I was once a rocket scientist and it wasn’t all it was cracked up to be.